clear all
set more off
set matsize 10000
tempfile ubigeo temp


local lista trust_police trust_army trust_church trust_newsp trust_rtv trust_parties trust_regi trust_jury trust_parliament trust_jne trust_onpe
local controls schooling male rural native married head 

use mergedpR, clear

egen voted_bn = rowmax(voted_blank voted_null)
recode voted_blank voted_null voted_bn (. = 0) if voted == 0

** LABELS
label var todoATR "Violence Exposure"

******* Results
drop if vote_matter == .
local lista vote_matter voted voted_blank voted_null voted_bn partN dem_works dem_matters1 gov_demo1 gov_auto1

local r = 1
foreach var in `lista' {
	replace `var' = `var' * 100 if `r' != 6
	reghdfe `var' todoATR `controls', abs(ubigeo1993 ubigeonac1993 i.ubigeo1993#i.ubigeonac1993 i.age i.year i.provid#c.trend) vce(cluster ubigeonac1993) keepsin
	est store reg`r'
	sum `var' if e(sample) == 1
	local y = r(mean)
	sum todoAT if e(sample) == 1 & todoAT
	local sd = r(mean)
	local n`r' = (round(((_b[todoAT]) * `sd') / `y', 0.0001)) * 100
	local ++r
}

noisily estout reg* ///
using Table_B3_0.tex, style(tex) c(b(fmt(2) star) se(par fmt(2))) ///
	label collabels(, none) eqlabels(none) mlabels(, none) varlabel(_cons " $ Control $ ") starlevel(* .1 ** .05 *** .01) ///
	keep(todoATR) ///
	order(todoATR) ///
	stats(r2 N N_clust, fmt(%9.2f %9.0g) labels(`"$ R ^{2}$"' "Observations" "Clusters")) replace 

texdoc init Table_B3.tex, replace
tex \begin{table}[h]
tex \caption{Impact of armed conflict (army, terrorist or rondas) on voting, participation and democratic beliefs}
tex \label{tab:table2_rondas}
tex \begin{center}
tex \resizebox{16cm}{!}{
tex \begin{tabular}{lcccccccccc} \hline \hline
tex \\
tex & Voting & \multicolumn{4}{c}{Voted in last elections}& Participation & \multicolumn{2}{c}{Democracy} & \multicolumn{2}{c}{Type of preferred gov.} \\
tex & matters & Yes & ...blank & ...null & ... blank or null & organisations & works & matters & Democracy & Autocracy \\
tex \hline
tex \input{Table_B3_0.tex}
tex &&&&&&&&&& \\
tex \multicolumn{8}{l}{\textbf{Effect of 1 s.d. change in conflict measure as \% of mean of dependent variable:}} \\
tex & `n1' & `n2' & `n3' & `n4' & `n5' & `n6' & `n7' & `n8' & `n9' & `n10' \\
tex &&&&&&&&&& \\
tex \hline
tex \multicolumn{11}{p{1.5\textwidth}}{\small \textbf{Notes:} ***, **, and * indicate statistical significance at the 1 percent, 5 percent and 10 percent level, respectively. Standard errors in parentheses, clustered at district of birth level. All models include fixed effects for district of birth, district of residence, the match of district of birth and residence, age, wave of the national household survey, and a province specific linear trend. In addition, all models control for gender, marital status, family size and indicator variables for head of household, urban residence and native language (proxy for indigenous population).The variable ``Violence Exposure'' represents the number of years of exposure to violence from either the government, rondas or from terrorist forces (the two potential \emph{perpetrators}) within the age interval 0 to 30 years old.}
tex \end{tabular}}
tex \end{center}
tex \end{table}
